MUMBAI, India, Jan. 9 -- Intellectual Property India has published a patent application (202521114766 A) filed by Symbiosis International, Pune, Maharashtra, on Nov. 20, 2025, for 'system and method for automated claim veracity classification using transformer-based multi-level truthfulness assessment.'
Inventor(s) include Shreyasee Shinde; Sneh Patel; Akshat Jain; Ojas Bodke; Aniket K. Shahade; Mayur Gaikwad; and Shruti Patil.
The application for the patent was published on Dec. 12, under issue no. 50/2025.
According to the abstract released by the Intellectual Property India: "A computer-implemented system for automated classification of claim veracity using multi-level truthfulness assessment is disclosed. The system comprises a processor and a memory storing executable instructions for data collection, preprocessing, tokenization, and classification modules. The data collection module aggregates claim data from multiple digital fact-checking sources. The preprocessing module consolidates heterogeneous truthfulness labels into standardized categories and performs data cleaning operations. The tokenization module generates token sequences with attention masks using transformer-based tokenizers. The classification module employs a bidirectional encoder model utilizing BERT architecture with multiple transformer layers to analyse token sequences in both forward and backward directions, generating contextual embeddings that capture semantic relationships within claim text. The system fine-tunes the model through iterative training processes optimizing parameters based on classification accuracy metrics."
Disclaimer: Curated by HT Syndication.
